Parallel Catmull-Rom Spline Interpolation Algorithm for Image Zooming Based on CUDA

TL;DR: The GPU WiSARD Vessel Tracker GWVT is proposed, a maritime vessel tracker that uses theWiSARD weightless neural network and implemented on a GPU, allowing other tracking techniques to be executed in parallel to rise the performance.

TL;DR: The heterogeneous implementation and evaluation of five algorithms to improve the resolution in magnetic resonance images, these are Nearest-Neighbor, Bilinear, Bicubic, Catmull-Rom and Lanczos are presented.

Abstract: The article is devoted to the development and application of an algorithm for analyzing graphomotor writing trajectories in digital educational systems using the Catmull-Rom spline. The problem of improving the accuracy and stability of processing user input received from touch devices (stylus or finger) in the context of digital writing learning is considered. The proposed method includes the stages of coordinate data collection, normalization, noise filtering, and trajectory smoothing. The Catmull-Roma spline is used to ensure the continuity of curves and preserve key motion features, which increases the accuracy of comparison with reference patterns. The paper presents a computational algorithm that implements an automatic assessment of the quality of letter writing based on a number of parameters: curvature, angular deviations, speed stability and the number of strokes. The algorithm is integrated into a digital diagnostic system capable of generating detailed reports, identifying common errors, and offering personalized tasks. The developed solution can be used in intelligent learning platforms, biometric authentication systems, neuropsychological diagnostics and computer forensics. The proposed approach demonstrates high adaptability to various input scenarios and provides the basis for building more complex machine learning systems focused on recognizing, analyzing, and generating handwritten text.